Replication: Incomes and Income Inequality in Stockholm, 1870–1970: Evidence from Micro Data

In the early 1900s, economic inequality was stark in the industrialized countries, but over the decades after the First World War, inequality fell in what has been labelled the Great Levelling. This paper studies the Swedish capital of Stockholm, using uniquely detailed micro data, to investigate mechanisms of the high inequality prevailing until about 1920, and the equalization thereafter. The investigation builds on a new dataset from the population register, comprising 38,022 randomly sampled Stockholm residents, using information about income for persons who earned enough to pay income tax, but also, for taxpayers and non-taxpayers alike, job title, age, and household composition, in the years 1870, 1880, 1900, 1910, 1920, 1940 and 1950. This is complemented with the Census of 1930 and a Statistics Sweden random sample for 1960 and 1970. The results show that incomes were very unequally distributed 1870–1920, with Gini coefficients of pre-tax income around sixty and with no statistically significant change. After 1920 inequality fell quite steadily for every benchmark year, to a Gini coefficient of to the mid-fifties in 1930, mid-forties in 1950, and the thirties in 1970. A drop in capital incomes contributed, and when looking at post-tax incomes the growth of progressive taxation after 1930 also contributed, but most of the high inequality up to 1920 and equalization after that depended on the distribution of labour income. We show that professional groups had huge income advantages over workers in the early 1900s, but that this advantage was drastically reduced after 1920 when working-class incomes grew rapidly. An important mechanism of income growth was the replacement of worse-paid jobs by better paid jobs. The share of domestic servants in the working class was reduced from 45 percent in 1870 to 10 percent in 1950. Given the decline of domestic service generally in the Western world in this period and the associated upgrading of the female work force, this has implications for understanding inequality also elsewhere. The results show the importance of structural change, beyond the Kuznetsian binary of agriculture–manufacturing, for understanding long-run economic inequality.